This guide provides a walkthrough for getting a Lobe TensorFlow model onto the Raspberry Pi 4.*

It is assumed you are starting with a Pi in a remote headless configuration already enabled with SSH, Remote Desktop or VNC, and WiFi access.

*Note that this procedure will work for a Pi 3, however it is unlikely the model will run successfully.

• Raspberry Pi 4
• SD Card with Raspbian (desktop version recommended)
• USB-C Power Supply (5.1V, 3A)
• Pi Camera
• Case (optional but recommended)
• Pi Camera Mount (optional but recommended)

• WinSCP (or other remote file transfer program)
• Remote Desktop or VNC

1. Carefully plug in your Pi Camera module. Instructions can be found here.
2. Power up your Pi and log in to the desktop view.
3. On the Pi, open a terminal window and create a directory where you want to store your Lobe model and code. For example, in the /home/pi folder:

   mkdir Lobe
   

4. Navigate into your Lobe directory, and install the lobe-python package:

   cd Lobe
   wget https://raw.githubusercontent.com/lobe/lobe-python/master/scripts/lobe-rpi-install.sh
   sudo ./lobe-rpi-install.sh
   

Note: you can also use a USB drive to transfer the files

1. On your PC or Mac, open WinSCP (or your preferred remote file transfer method) and connect to the Pi.

2. On the Pi side navigate to the Lobe directory and create a new folder called "model".

3. Select and the Lobe TensorFlow model files on your PC or Mac into the model folder on the Pi.

   Note: You should have two files -- "saved_model.pb" and "signature.json" -- as well as a folder named "variables" with two files inside.

1. On the Pi, open the rpi_quickstart.py file with your favorite Python editor.
2. In line 4, update the model path to point to the folder with your Lobe TF model files. Hint: if you created the directory with the same name as above you don't need to change anything.
3. Run your python program with Python 3: ''' python3 rpi_quickstart.py '''

1. Check that the Pi power light is bright red. A dim red light indicates insufficient power.
2. Be sure you're installing and running your TF code using Python 3.
3. If you're using multiple components, be sure they are all connected to the same common ground.